CT-guided High-Dose-Rate Brachytherapy versus Transarterial Chemoembolization in Patients with Unresectable Hepatocellular Carcinoma.

Background CT-guided high-dose-rate (HDR) brachytherapy (hereafter, HDR brachytherapy) has been shown to be safe and effective for patients with unresectable hepatocellular carcinoma (HCC), but studies comparing this therapy with other local-regional therapies are scarce. Purpose To compare patient outcomes of HDR brachytherapy and transarterial chemoembolization (TACE) in patients with unresectable HCC. Materials and Methods This multi-institutional retrospective study included consecutive treatment-naive adult patients with unresectable HCC who underwent either HDR brachytherapy or TACE between January 2010 and December 2022. Overall survival (OS) and progression-free survival (PFS) were compared between patients matched for clinical and tumor characteristics by propensity score matching. Not all patients who underwent TACE had PFS available; thus, a different set of patients was used for PFS and OS analysis for this treatment. Hazard ratios (HRs) were calculated from Kaplan-Meier survival curves. Results After propensity matching, 150 patients who underwent HDR brachytherapy (median age, 71 years [IQR, 63-77 years]; 117 males) and 150 patients who underwent TACE (OS analysis median age, 70 years [IQR, 63-77 years]; 119 male; PFS analysis median age, 68 years [IQR: 63-76 years]; 119 male) were analyzed. Hazard of death was higher in the TACE versus HDR brachytherapy group (HR, 4.04; P < .001). Median estimated PFS was 32.8 months (95% CI: 12.5, 58.7) in the HDR brachytherapy group and 11.6 months (95% CI: 4.9, 22.7) in the TACE group. Hazard of disease progression was higher in the TACE versus HDR brachytherapy group (HR, 2.23; P < .001). Conclusion In selected treatment-naive patients with unresectable HCC, treatment with CT-guided HDR brachytherapy led to improved OS and PFS compared with TACE. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Chapiro in this issue.

Towards reproducible radiomics research: introduction of a database for radiomics studies.

OBJECTIVES: To investigate the model-, code-, and data-sharing practices in the current radiomics research landscape and to introduce a radiomics research database. METHODS: A total of 1254 articles published between January 1, 2021, and December 31, 2022, in leading radiology journals (European Radiology, European Journal of Radiology, Radiology, Radiology: Artificial Intelligence, Radiology: Cardiothoracic Imaging, Radiology: Imaging Cancer) were retrospectively screened, and 257 original research articles were included in this study. The categorical variables were compared using Fisher's exact tests or chi-square test and numerical variables using Student's t test with relation to the year of publication. RESULTS: Half of the articles (128 of 257) shared the model by either including the final model formula or reporting the coefficients of selected radiomics features. A total of 73 (28%) models were validated on an external independent dataset. Only 16 (6%) articles shared the data or used publicly available open datasets. Similarly, only 20 (7%) of the articles shared the code. A total of 7 (3%) articles both shared code and data. All collected data in this study is presented in a radiomics research database (RadBase) and could be accessed at https://github.com/EuSoMII/RadBase . CONCLUSION: According to the results of this study, the majority of published radiomics models were not technically reproducible since they shared neither model nor code and data. There is still room for improvement in carrying out reproducible and open research in the field of radiomics. CLINICAL RELEVANCE STATEMENT: To date, the reproducibility of radiomics research and open science practices within the radiomics research community are still very low. Ensuring reproducible radiomics research with model-, code-, and data-sharing practices will facilitate faster clinical translation. KEY POINTS: • There is a discrepancy between the number of published radiomics papers and the clinical implementation of these published radiomics models. • The main obstacle to clinical implementation is the lack of model-, code-, and data-sharing practices. • In order to translate radiomics research into clinical practice, the radiomics research community should adopt open science practices.

Reproducibility of radiomics quality score: an intra- and inter-rater reliability study.

OBJECTIVES: To investigate the intra- and inter-rater reliability of the total radiomics quality score (RQS) and the reproducibility of individual RQS items' score in a large multireader study. METHODS: Nine raters with different backgrounds were randomly assigned to three groups based on their proficiency with RQS utilization: Groups 1 and 2 represented the inter-rater reliability groups with or without prior training in RQS, respectively; group 3 represented the intra-rater reliability group. Thirty-three original research papers on radiomics were evaluated by raters of groups 1 and 2. Of the 33 papers, 17 were evaluated twice with an interval of 1 month by raters of group 3. Intraclass coefficient (ICC) for continuous variables, and Fleiss' and Cohen's kappa (k) statistics for categorical variables were used. RESULTS: The inter-rater reliability was poor to moderate for total RQS (ICC 0.30-055, p < 0.001) and very low to good for item's reproducibility (k - 0.12 to 0.75) within groups 1 and 2 for both inexperienced and experienced raters. The intra-rater reliability for total RQS was moderate for the less experienced rater (ICC 0.522, p = 0.009), whereas experienced raters showed excellent intra-rater reliability (ICC 0.91-0.99, p < 0.001) between the first and second read. Intra-rater reliability on RQS items' score reproducibility was higher and most of the items had moderate to good intra-rater reliability (k - 0.40 to 1). CONCLUSIONS: Reproducibility of the total RQS and the score of individual RQS items is low. There is a need for a robust and reproducible assessment method to assess the quality of radiomics research. CLINICAL RELEVANCE STATEMENT: There is a need for reproducible scoring systems to improve quality of radiomics research and consecutively close the translational gap between research and clinical implementation. KEY POINTS: • Radiomics quality score has been widely used for the evaluation of radiomics studies. • Although the intra-rater reliability was moderate to excellent, intra- and inter-rater reliability of total score and point-by-point scores were low with radiomics quality score. • A robust, easy-to-use scoring system is needed for the evaluation of radiomics research.

Does Dual-Energy Computed Tomography Material Decomposition Improve Radiomics Capability to Predict Survival in Head and Neck Squamous Cell Carcinoma Patients? A Preliminary Investigation.

OBJECTIVE: Our study objective was to explore the additional value of dual-energy CT (DECT) material decomposition for squamous cell carcinoma of the head and neck (SCCHN) survival prognostication. METHODS: A group of 50 SCCHN patients (male, 37; female, 13; mean age, 63.6 ± 10.82 years) with baseline head and neck DECT between September 2014 and August 2020 were retrospectively included. Primary tumors were segmented, radiomics features were extracted, and DECT material decomposition was performed. We used independent train and validation datasets with cross-validation and 100 independent iterations to identify prognostic signatures applying elastic net (EN) and random survival forest (RSF). Features were ranked and intercorrelated according to their prognostic importance. We benchmarked the models against clinical parameters. Intraclass correlation coefficients were used to analyze the interreader variation. RESULTS: The exclusively radiomics-trained models achieved similar ( P = 0.947) prognostic performance of area under the curve (AUC) = 0.784 (95% confidence interval [CI], 0.775-0.812) (EN) and AUC = 0.785 (95% CI, 0.759-0.812) (RSF). The additional application of DECT material decomposition did not improve the model's performance (EN, P = 0.594; RSF, P = 0.198). In the clinical benchmark, the top averaged AUC value of 0.643 (95% CI, 0.611-0.675) was inferior to the quantitative imaging-biomarker models ( P < 0.001). A combined imaging and clinical model did not improve the imaging-based models ( P > 0.101). Shape features revealed high prognostic importance. CONCLUSIONS: Radiomics AI applications may be used for SCCHN survival prognostication, but the spectral information of DECT material decomposition did not improve the model's performance in our preliminary investigation.

Comparison of 96-kV and 120-kV cone-beam CT for the assessment of cochlear implants.

BACKGROUND: To compare the diagnostic value of 120-kV with conventional 96-kV Cone-Beam CT (CBCT) of the temporal bone after cochlear implant (CI) surgery. METHODS: This retrospective study included CBCT scans after CI surgery between 06/17 and 01/18. CBCT allowed examinations with 96-kV or 120-kV; other parameters were the same. Two radiologists independently evaluated following criteria on 5-point Likert scales: osseous spiral lamina, inner and outer cochlear wall, semi-circular canals, mastoid trabecular structure, overall image quality, metal and motion artefacts, depiction of intracochlear electrode position and visualisation of single electrode contacts. Effective radiation dose was assessed. RESULTS: Seventy-five patients (females, n = 39 [52.0%], mean age, 55.8 ± 16.5 years) were scanned with 96-kV (n = 32, 42.7%) and 120-kV (n = 43, 57.3%) protocols including CI models from three vendors (vendor A n = 7; vendor B n = 43; vendor C n = 25). Overall image quality, depiction of anatomical structures, and electrode position were rated significantly better in 120-kV images compared to 96-kV (all p < = 0.018). Anatomical structures and electrode position were rated significantly better in 120-kV CBCT for CI models from vendor A and C, while 120-kV did not provide improved image quality in CI models from vendor B. Radiation doses were significantly higher for 120-kV scans compared to 96-kV (0.15 vs. 0.08 mSv, p < 0.001). CONCLUSIONS: 120-kV and 96-kV CBCT provide good diagnostic images for the postoperative CI evaluation. While 120-kV showed improved depiction of temporal bone and CI electrode position compared to 96-kV in most CI models, the 120-kV protocol should be chosen wisely due to a substantially higher radiation exposure.

Automation Bias in Mammography: The Impact of Artificial Intelligence BI-RADS Suggestions on Reader Performance.

Background Automation bias (the propensity for humans to favor suggestions from automated decision-making systems) is a known source of error in human-machine interactions, but its implications regarding artificial intelligence (AI)-aided mammography reading are unknown. Purpose To determine how automation bias can affect inexperienced, moderately experienced, and very experienced radiologists when reading mammograms with the aid of an artificial intelligence (AI) system. Materials and Methods In this prospective experiment, 27 radiologists read 50 mammograms and provided their Breast Imaging Reporting and Data System (BI-RADS) assessment assisted by a purported AI system. Mammograms were obtained between January 2017 and December 2019 and were presented in two randomized sets. The first was a training set of 10 mammograms, with the correct BI-RADS category suggested by the AI system. The second was a set of 40 mammograms in which an incorrect BI-RADS category was suggested for 12 mammograms. Reader performance, degree of bias in BI-RADS scoring, perceived accuracy of the AI system, and reader confidence in their own BI-RADS ratings were assessed using analysis of variance (ANOVA) and repeated-measures ANOVA followed by post hoc tests and Kruskal-Wallis tests followed by the Dunn post hoc test. Results The percentage of correctly rated mammograms by inexperienced (mean, 79.7% ± 11.7 [SD] vs 19.8% ± 14.0; P < .001; r = 0.93), moderately experienced (mean, 81.3% ± 10.1 vs 24.8% ± 11.6; P < .001; r = 0.96), and very experienced (mean, 82.3% ± 4.2 vs 45.5% ± 9.1; P = .003; r = 0.97) radiologists was significantly impacted by the correctness of the AI prediction of BI-RADS category. Inexperienced radiologists were significantly more likely to follow the suggestions of the purported AI when it incorrectly suggested a higher BI-RADS category than the actual ground truth compared with both moderately (mean degree of bias, 4.0 ± 1.8 vs 2.4 ± 1.5; P = .044; r = 0.46) and very (mean degree of bias, 4.0 ± 1.8 vs 1.2 ± 0.8; P = .009; r = 0.65) experienced readers. Conclusion The results show that inexperienced, moderately experienced, and very experienced radiologists reading mammograms are prone to automation bias when being supported by an AI-based system. This and other effects of human and machine interaction must be considered to ensure safe deployment and accurate diagnostic performance when combining human readers and AI. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Baltzer in this issue.

Dual-Energy CT-based Opportunistic Volumetric Bone Mineral Density Assessment of the Distal Radius.

Background In patients with distal radius fractures (DRFs), low bone mineral density (BMD) is associated with bone substitute use during surgery and bone nonunion, but BMD information is not regularly available. Purpose To evaluate the feasibility of dual-energy CT (DECT)-based BMD assessment from routine examinations in the distal radius and the relationship between the obtained BMD values, the occurrence of DRFs, bone nonunion, and use of surgical bone substitute. Materials and Methods Scans in patients who underwent routine dual-source DECT in the distal radius between January 2016 and December 2021 were retrospectively acquired. Phantomless BMD assessment was performed using the delineated trabecular bone of a nonfractured segment of the distal radius and both DECT image series. CT images and health records were examined to determine fracture severity, surgical management, and the occurrence of bone nonunion. Associations of BMD with the occurrence of DRFs, bone nonunion, and bone substitute use at surgical treatment were examined with generalized additive models and receiver operating characteristic analysis. Results This study included 263 patients (median age, 52 years; IQR, 36-64 years; 132 female patients), of whom 192 were diagnosed with fractures. Mean volumetric BMD was lower in patients who sustained a DRF (93.9 mg/cm3 vs 135.4 mg/cm3; P < .001), required bone substitutes (79.6 mg/cm3 vs 95.5 mg/cm3; P < .001), and developed bone nonunion (71.1 mg/cm3 vs 96.5 mg/cm3; P < .001). Receiver operating characteristic curve analysis identified these patients with an area under the curve of 0.71-0.91 (P < .001). Lower BMD increased the risk to sustain DRFs, develop bone nonunion, and receive bone substitutes at surgery (P < .001). Conclusion DECT-based BMD assessment at routine examinations is feasible and could help predict surgical bone substitute use and the occurrence of bone nonunion in patients with DRFs. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Carrino in this issue.

Unveiling the diagnostic enigma of D-dimer testing in cancer patients: Current evidence and areas of application.

BACKGROUND: Cancer is a well-known risk factor for venous thromboembolism (VTE). A combined strategy of D-dimer testing and clinical pre-test probability is usually used to exclude VTE. However, its effectiveness is diminished in cancer patients due to reduced specificity, ultimately leading to a decreased clinical utility. This review article seeks to provide a comprehensive summary of how to interpret D-dimer testing in cancer patients. METHODS: In accordance with PRISMA standards, literature pertaining to the diagnostic and prognostic significance of D-dimer testing in cancer patients was carefully chosen from reputable sources such as PubMed and the Cochrane databases. RESULTS: D-dimers have not only a diagnostic value in ruling out VTE but can also serve as an aid for rule-in if their values exceed 10-times the upper limit of normal. This threshold allows a diagnosis of VTE in cancer patients with a positive predictive value of more than 80%. Moreover, elevated D-dimers carry important prognostic information and are associated with VTE reoccurrence. A gradual increase in risk for all-cause death suggests that VTE is also an indicator of biologically more aggressive cancer types and advanced cancer stages. Considering the lack of standardization for D-dimer assays, it is essential for clinicians to carefully consider the variations in assay performance and the specific test characteristics of their institution. CONCLUSIONS: Standardizing D-dimer assays and developing modified pretest probability models specifically for cancer patients, along with adjusted cut-off values for D-dimer testing, could significantly enhance the accuracy and effectiveness of VTE diagnosis in this population.

Trends and statistics of artificial intelligence and radiomics research in Radiology, Nuclear Medicine, and Medical Imaging: bibliometric analysis.

OBJECTIVE: To conduct a comprehensive bibliometric analysis of artificial intelligence (AI) and its subfields as well as radiomics in Radiology, Nuclear Medicine, and Medical Imaging (RNMMI). METHODS: Web of Science was queried for relevant publications in RNMMI and medicine along with their associated data from 2000 to 2021. Bibliometric techniques utilised were co-occurrence, co-authorship, citation burst, and thematic evolution analyses. Growth rate and doubling time were also estimated using log-linear regression analyses. RESULTS: According to the number of publications, RNMMI (11,209; 19.8%) was the most prominent category in medicine (56,734). USA (44.6%) and China (23.1%) were the two most productive and collaborative countries. USA and Germany experienced the strongest citation bursts. Thematic evolution has recently exhibited a significant shift toward deep learning. In all analyses, the annual number of publications and citations demonstrated exponential growth, with deep learning-based publications exhibiting the most prominent growth pattern. Estimated continuous growth rate, annual growth rate, and doubling time of the AI and machine learning publications in RNMMI were 26.1% (95% confidence interval [CI], 12.0-40.2%), 29.8% (95% CI, 12.7-49.5%), and 2.7 years (95% CI, 1.7-5.8), respectively. In the sensitivity analysis using data from the last 5 and 10 years, these estimates ranged from 47.6 to 51.1%, 61.0 to 66.7%, and 1.4 to 1.5 years. CONCLUSION: This study provides an overview of AI and radiomics research conducted mainly in RNMMI. These results may assist researchers, practitioners, policymakers, and organisations in gaining a better understanding of both the evolution of these fields and the importance of supporting (e.g., financial) these research activities. KEY POINTS: • In terms of the number of publications on AI and ML, Radiology, Nuclear Medicine, and Medical Imaging was the most prominent category compared to the other categories related to medicine (e.g., Health Policy & Services, Surgery). • All evaluated analyses (i.e., AI, its subfields, and radiomics), based on the annual number of publications and citations, demonstrated exponential growth, with decreasing doubling time, which indicates increasing interest from researchers, journals, and, in turn, the medical imaging community. • The most prominent growth pattern was observed in deep learning-based publications. However, the further thematic analysis demonstrated that deep learning has been underdeveloped but highly relevant to the medical imaging community.

Low bone mineral density is a prognostic factor for elderly patients with HCC undergoing TACE: results from a multicenter study.

OBJECTIVES: Low bone mineral density (BMD) was recently identified as a novel risk factor for patients with hepatocellular carcinoma (HCC). In this multicenter study, we aimed to validate the role of BMD as a prognostic factor for patients with HCC undergoing transarterial chemoembolization (TACE). METHODS: This retrospective multicenter trial included 908 treatment-naïve patients with HCC who were undergoing TACE as a first-line treatment, at six tertiary care centers, between 2010 and 2020. BMD was assessed by measuring the mean Hounsfield units (HUs) in the midvertebral core of the 11th thoracic vertebra, on contrast-enhanced computer tomography performed before treatment. We assessed the influence of BMD on median overall survival (OS) and performed multivariate analysis including established estimates for survival. RESULTS: The median BMD was 145 HU (IQR, 115-175 HU). Patients with a high BMD (≥ 114 HU) had a median OS of 22.2 months, while patients with a low BMD (< 114 HU) had a lower median OS of only 16.2 months (p < .001). Besides albumin, bilirubin, tumor number, and tumor diameter, BMD remained an independent prognostic factor in multivariate analysis. CONCLUSIONS: BMD is an independent predictive factor for survival in elderly patients with HCC undergoing TACE. The integration of BMD into novel scoring systems could potentially improve survival prediction and clinical decision-making. KEY POINTS: • Bone mineral density can be easily assessed in routinely acquired pre-interventional computed tomography scans. • Bone mineral density is an independent predictive factor for survival in elderly patients with HCC undergoing TACE. • Thus, bone mineral density is a novel imaging biomarker for prognosis prediction in elderly patients with HCC undergoing TACE.

Fully automated AI-based splenic segmentation for predicting survival and estimating the risk of hepatic decompensation in TACE patients with HCC.

OBJECTIVES: Splenic volume (SV) was proposed as a relevant prognostic factor for patients with hepatocellular carcinoma (HCC). We trained a deep-learning algorithm to fully automatically assess SV based on computed tomography (CT) scans. Then, we investigated SV as a prognostic factor for patients with HCC undergoing transarterial chemoembolization (TACE). METHODS: This retrospective study included 327 treatment-naïve patients with HCC undergoing initial TACE at our tertiary care center between 2010 and 2020. A convolutional neural network was trained and validated on the first 100 consecutive cases for spleen segmentation. Then, we used the algorithm to evaluate SV in all 327 patients. Subsequently, we evaluated correlations between SV and survival as well as the risk of hepatic decompensation during TACE. RESULTS: The algorithm showed Sørensen Dice Scores of 0.96 during both training and validation. In the remaining 227 patients assessed with the algorithm, spleen segmentation was visually approved in 223 patients (98.2%) and failed in four patients (1.8%), which required manual re-assessments. Mean SV was 551 ml. Survival was significantly lower in patients with high SV (10.9 months), compared to low SV (22.0 months, p = 0.001). In contrast, overall survival was not significantly predicted by axial and craniocaudal spleen diameter. Furthermore, patients with a hepatic decompensation after TACE had significantly higher SV (p < 0.001). CONCLUSION: Automated SV assessments showed superior survival predictions in patients with HCC undergoing TACE compared to two-dimensional spleen size estimates and identified patients at risk of hepatic decompensation. Thus, SV could serve as an automatically available, currently underappreciated imaging biomarker. KEY POINTS: • Splenic volume is a relevant prognostic factor for prediction of survival in patients with HCC undergoing TACE, and should be preferred over two-dimensional surrogates for splenic size. • Besides overall survival, progression-free survival and hepatic decompensation were significantly associated with splenic volume, making splenic volume a currently underappreciated prognostic factor prior to TACE. • Splenic volume can be fully automatically assessed using deep-learning methods; thus, it is a promising imaging biomarker easily integrable into daily radiological routine.

Two-dimensional CT measurements enable assessment of body composition on head and neck CT.

OBJECTIVES: The aim of this study was to evaluate whether simple 2D measurements in axial slices of head and neck CT examinations correlate with generally established measurements of body composition in abdominal CT at the height of the third lumbar vertebra and thus allow for an estimation of muscle and fat masses. METHODS: One hundred twenty-two patients who underwent concurrent CT of the head and neck and the abdomen between July 2016 and July 2020 were retrospectively included. For a subset of 30 patients, additional bioelectrical impedance analysis (BIA) was available. Areas of paraspinal muscles at the height of the third (C3) and fifth cervical vertebrae (C5) as well as the total cross-sectional area at the height of C3 and at the submandibular level were correlated with the results of abdominal measurements and BIA. Furthermore, intra- and interreader variabilities of all measurements were assessed. RESULTS: Regarding adipose tissue, good correlations were found between the total cross-sectional area of the patient's body at the submandibular level and at the height of C3 between both abdominal measurements and BIA results (r = 0.8-0.92; all p < 0.001). Regarding muscle, the total paraspinal muscle area at the height of C3 and C5 showed strong correlations with abdominal measurements and moderate to strong correlations with BIA results (r = 0.44-0.80; all p < 0.001), with the muscle area on C5 yielding slightly higher correlations. CONCLUSIONS: Body composition information can be obtained with comparable reliability from head and neck CT using simple biplanar measurements as from abdominal CT. KEY POINTS: • The total paraspinal muscle area at the height of C3 and C5 correlates strongly with abdominal muscle mass. • The total cross-sectional area at the submandibular level and at the height of C3 shows good correlations with abdominal fat mass. • The described measurements facilitate a rapid, opportunistic assessment of relevant body composition parameters.

Commercial AI solutions in detecting COVID-19 pneumonia in chest CT: not yet ready for clinical implementation?

OBJECTIVES: In response to the COVID-19 pandemic, many researchers have developed artificial intelligence (AI) tools to differentiate COVID-19 pneumonia from other conditions in chest CT. However, in many cases, performance has not been clinically validated. The aim of this study was to evaluate the performance of commercial AI solutions in differentiating COVID-19 pneumonia from other lung conditions. METHODS: Four commercial AI solutions were evaluated on a dual-center clinical dataset consisting of 500 CT studies; COVID-19 pneumonia was microbiologically proven in 50 of these. Sensitivity, specificity, positive and negative predictive values, and AUC were calculated. In a subgroup analysis, the performance of the AI solutions in differentiating COVID-19 pneumonia from other conditions was evaluated in CT studies with ground-glass opacities (GGOs). RESULTS: Sensitivity and specificity ranges were 62-96% and 31-80%, respectively. Negative and positive predictive values ranged between 82-99% and 19-25%, respectively. AUC was in the range 0.54-0.79. In CT studies with GGO, sensitivity remained unchanged. However, specificity was lower, and ranged between 15 and 53%. AUC for studies with GGO was in the range 0.54-0.69. CONCLUSIONS: This study highlights the variable specificity and low positive predictive value of AI solutions in diagnosing COVID-19 pneumonia in chest CT. However, one solution yielded acceptable values for sensitivity. Thus, with further improvement, commercial AI solutions currently under development have the potential to be integrated as alert tools in clinical routine workflow. Randomized trials are needed to assess the true benefits and also potential harms of the use of AI in image analysis. KEY POINTS: • Commercial AI solutions achieved a sensitivity and specificity ranging from 62 to 96% and from 31 to 80%, respectively, in identifying patients suspicious for COVID-19 in a clinical dataset. • Sensitivity remained within the same range, while specificity was even lower in subgroup analysis of CT studies with ground-glass opacities, and interrater agreement between the commercial AI solutions was minimal to nonexistent. • Thus, commercial AI solutions have the potential to be integrated as alert tools for the detection of patients with lung changes suspicious for COVID-19 pneumonia in a clinical routine workflow, if further improvement is made.

Artificial intelligence reporting guidelines: what the pediatric radiologist needs to know.

There has been an exponential rise in artificial intelligence (AI) research in imaging in recent years. While the dissemination of study data that has the potential to improve clinical practice is welcomed, the level of detail included in early AI research reporting has been highly variable and inconsistent, particularly when compared to more traditional clinical research. However, inclusion checklists are now commonly available and accessible to those writing or reviewing clinical research papers. AI-specific reporting guidelines also exist and include distinct requirements, but these can be daunting for radiologists new to the field. Given that pediatric radiology is a specialty faced with workforce shortages and an ever-increasing workload, AI could help by offering solutions to time-consuming tasks, thereby improving workflow efficiency and democratizing access to specialist opinion. As a result, pediatric radiologists are expected to be increasingly leading and contributing to AI imaging research, and researchers and clinicians alike should feel confident that the findings reported are presented in a transparent way, with sufficient detail to understand how they apply to wider clinical practice. In this review, we describe two of the most clinically relevant and available reporting guidelines to help increase awareness and engage the pediatric radiologist in conducting AI imaging research. This guide should also be useful for those reading and reviewing AI imaging research and as a checklist with examples of what to expect.

Radiomics for prediction of survival in lower-grade gliomas-it's time to move beyond the crystal ball.

KEY POINTS: • Radiomics might help predict survival of patients with lower-grade gliomas.• Several different models using different radiomics features have been proposed with only little overlap in included features.• Prospective trials and validation studies are needed to establish which models offer clinical benefit and which do not.

A decade of radiomics research: are images really data or just patterns in the noise?

KEY POINTS: • Although radiomics is potentially a promising approach to analyze medical image data, many pitfalls need to be considered to avoid a reproducibility crisis.• There is a translation gap in radiomics research, with many studies being published but so far little to no translation into clinical practice.• Going forward, more studies with higher levels of evidence are needed, ideally also focusing on prospective studies with relevant clinical impact.

Practical applications of deep learning: classifying the most common categories of plain radiographs in a PACS using a neural network.

OBJECTIVES: The goal of the present study was to classify the most common types of plain radiographs using a neural network and to validate the network's performance on internal and external data. Such a network could help improve various radiological workflows. METHODS: All radiographs from the year 2017 (n = 71,274) acquired at our institution were retrieved from the PACS. The 30 largest categories (n = 58,219, 81.7% of all radiographs performed in 2017) were used to develop and validate a neural network (MobileNet v1.0) using transfer learning. Image categories were extracted from DICOM metadata (study and image description) and mapped to the WHO manual of diagnostic imaging. As an independent, external validation set, we used images from other institutions that had been stored in our PACS (n = 5324). RESULTS: In the internal validation, the overall accuracy of the model was 90.3% (95%CI: 89.2-91.3%), whereas, for the external validation set, the overall accuracy was 94.0% (95%CI: 93.3-94.6%). CONCLUSIONS: Using data from one single institution, we were able to classify the most common categories of radiographs with a neural network. The network showed good generalizability on the external validation set and could be used to automatically organize a PACS, preselect radiographs so that they can be routed to more specialized networks for abnormality detection or help with other parts of the radiological workflow (e.g., automated hanging protocols; check if ordered image and performed image are the same). The final AI algorithm is publicly available for evaluation and extension. KEY POINTS: • Data from one single institution can be used to train a neural network for the correct detection of the 30 most common categories of plain radiographs. • The trained model achieved a high accuracy for the majority of categories and showed good generalizability to images from other institutions. • The neural network is made publicly available and can be used to automatically organize a PACS or to preselect radiographs so that they can be routed to more specialized neural networks for abnormality detection.

To buy or not to buy-evaluating commercial AI solutions in radiology (the ECLAIR guidelines).

Artificial intelligence (AI) has made impressive progress over the past few years, including many applications in medical imaging. Numerous commercial solutions based on AI techniques are now available for sale, forcing radiology practices to learn how to properly assess these tools. While several guidelines describing good practices for conducting and reporting AI-based research in medicine and radiology have been published, fewer efforts have focused on recommendations addressing the key questions to consider when critically assessing AI solutions before purchase. Commercial AI solutions are typically complicated software products, for the evaluation of which many factors are to be considered. In this work, authors from academia and industry have joined efforts to propose a practical framework that will help stakeholders evaluate commercial AI solutions in radiology (the ECLAIR guidelines) and reach an informed decision. Topics to consider in the evaluation include the relevance of the solution from the point of view of each stakeholder, issues regarding performance and validation, usability and integration, regulatory and legal aspects, and financial and support services. KEY POINTS: • Numerous commercial solutions based on artificial intelligence techniques are now available for sale, and radiology practices have to learn how to properly assess these tools. • We propose a framework focusing on practical points to consider when assessing an AI solution in medical imaging, allowing all stakeholders to conduct relevant discussions with manufacturers and reach an informed decision as to whether to purchase an AI commercial solution for imaging applications. • Topics to consider in the evaluation include the relevance of the solution from the point of view of each stakeholder, issues regarding performance and validation, usability and integration, regulatory and legal aspects, and financial and support services.

Pneumatosis intestinalis and porto-mesenteric venous gas: a multicenter study.

BACKGROUND: Estimating the prognosis of patients with pneumatosis intestinalis (PI) and porto-mesenteric venous gas (PMVG) can be challenging. The purpose of this study was to refine prognostication to improve decision making in daily clinical routine. METHODS: A total of 290 patients with confirmed PI were included in the final analysis. The presence of PMVG and mortality (90d follow-up) were evaluated with regard to the influence of possible risk factors. Furthermore, a linear estimation model was devised combining significant parameters to calculate accuracies for predicting death in patients undergoing surgery by means of a defined operation point (ROC-analysis). RESULTS: Overall, 90d mortality was 55.2% (160/290). In patients with PI only, mortality was 46.5% (78/168) and increased significantly to 67.2% (82/122) in combination with PMVG (median survival: PI: 58d vs. PI and PMVG: 41d; p < 0.001). In the entire patient group, 53.5% (155/290) were treated surgically with a 90d mortality of 58.8% (91/155) in this latter group, while 90d mortality was 51.1% (69/135) in patients treated conservatively. In the patients who survived > 90d treated conservatively (24.9% of the entire collective; 72/290) PMVG/PI was defined as "benign"/reversible. PMVG, COPD, sepsis and a low platelet count were found to correlate with a worse prognosis helping to identify patients who might not profit from surgery, in this context our calculation model reaches accuracies of 97% specificity, 20% sensitivity, 90% PPV and 45% NPV. CONCLUSION: Although PI is associated with high morbidity and mortality, "benign causes" are common. However, in concomitant PMVG, mortality rates increase significantly. Our mathematical model could serve as a decision support tool to identify patients who are least likely to benefit from surgery, and to potentially reduce overtreatment in this subset of patients.

Use and Control of Artificial Intelligence in Patients Across the Medical Workflow: Single-Center Questionnaire Study of Patient Perspectives.

BACKGROUND: Artificial intelligence (AI) is gaining increasing importance in many medical specialties, yet data on patients' opinions on the use of AI in medicine are scarce. OBJECTIVE: This study aimed to investigate patients' opinions on the use of AI in different aspects of the medical workflow and the level of control and supervision under which they would deem the application of AI in medicine acceptable. METHODS: Patients scheduled for computed tomography or magnetic resonance imaging voluntarily participated in an anonymized questionnaire between February 10, 2020, and May 24, 2020. Patient information, confidence in physicians vs AI in different clinical tasks, opinions on the control of AI, preference in cases of disagreement between AI and physicians, and acceptance of the use of AI for diagnosing and treating diseases of different severity were recorded. RESULTS: In total, 229 patients participated. Patients favored physicians over AI for all clinical tasks except for treatment planning based on current scientific evidence. In case of disagreement between physicians and AI regarding diagnosis and treatment planning, most patients preferred the physician's opinion to AI (96.2% [153/159] vs 3.8% [6/159] and 94.8% [146/154] vs 5.2% [8/154], respectively; P=.001). AI supervised by a physician was considered more acceptable than AI without physician supervision at diagnosis (confidence rating 3.90 [SD 1.20] vs 1.64 [SD 1.03], respectively; P=.001) and therapy (3.77 [SD 1.18] vs 1.57 [SD 0.96], respectively; P=.001). CONCLUSIONS: Patients favored physicians over AI in most clinical tasks and strongly preferred an application of AI with physician supervision. However, patients acknowledged that AI could help physicians integrate the most recent scientific evidence into medical care. Application of AI in medicine should be disclosed and controlled to protect patient interests and meet ethical standards.